microvm

Toolkit for Seamlessly Enabling RL Training on Any Agent with Bedrock AgentCore.

Secure code execution in isolated lightweight VMs (QEMU microVMs). Python library for running untrusted Python, JavaScript, and shell code with 9-layer security isolation.

🦞 OpenClaw Pool — multi-tenant AI agent platform on AWS EC2 Firecracker microVMs with per-tenant isolation, auto-scaling, and web management console.

Security-posture-first list of sandboxes for running AI coding agents' code, ranked by isolation, egress control and secrets handling

Multi-tenant AI agents on AWS Lambda MicroVMs — one Firecracker VM per tenant with near-zero idle cost, snapshot resume, auto lifecycle, and EFS-persisted state.

Run code in isolated cloud containers. ConTree gives secure sandboxed execution environments with full root access, network, and persistent images.

Minimal, production-ready container operating systems by Miget — built on Ubuntu and Alpine with selectable runtimes (Docker, Podman) for secure, cloud-native workloads.

🦞 OpenClaw Pool — multi-tenant AI agent platform on AWS EC2 Firecracker microVMs with per-tenant isolation, auto-scaling, and web management console.

strangeClaw is a minimal, self-hosted autonomous AI agent with full Firecracker sandbox isolation support.

Command-line client for the ConTree sandboxing platform: VM-isolated secure sandboxes with git-like branching for AI agents and developers

Easy, fast, isolated microVM sandboxes for AI agents and untrusted code on AWS Lambda MicroVMs — Python SDK, asb CLI, and MCP server.

Python SDK for Declaw — security-first sandboxing for AI agents

Runnable Modal VM Sandbox examples — run Docker, docker compose, docker build, and filesystem snapshots on Modal's microVM runtime (vm_runtime). Reproducible Modal VM examples.

MicroVM specialized container build system in Python

Firecracker microVM sandboxes for AI agents — boot Linux in milliseconds, run code, write files, expose ports over HTTPS, then pause a session with memory intact and resume it later. Templates: base and code-interpreter-v1. Python and TypeScript SDK for Lizard Sandboxes.

Run code in isolated cloud containers. ConTree gives AI agents secure sandboxed execution environments with full root access, network, and persistent images.

Apple Silicon native AI agent sandbox — three-tier isolated code execution with MCP server. pip install silicon-sandbox

A disposable computer you and your AI share, streamed into your chat.

Dify plugin for Declaw — secure AI sandboxes with network policies, PII scanning, injection defense, and audit logging

Legacy NixOS microVM launcher for the Hebrah connection sidecar (superseded by hebrah-vm-templates)